sPHENIX tracking discussion

[Anthony Frawley <afrawley AT fsu.edu>]

Hello Jamie,

I will have a go at one of your questions tonight.

* Some of the current momentum resolution / upsilon resolution plots look like that program is lost (or never was).    It is possible to produce a figure of (1) charged particle momentum resolution versus p up to 50 GeV, and (2) upsilon 3-state mass resolution for the following equivalent rates: 1 kHz, 10 kHz, 25 kHz, 50 kHz.     At what rate do the effects of degraded resolution  really kick in?

The tracking software chain has been completely replaced in the last year (mostly in the last 6 months). We now have the tools we want in place, but we are still in the process of finalizing, testing and tuning things. Performance is changing on a weekly, sometimes daily, basis. In particular, we are just beginning to seriously grapple with high occupancy performance. The major problems we have now with momentum resolution and tracking efficiency are related to clustering and track seeding in the TPC. The track seeding is still very much under active development. Replacing our too-simple TPC clustering code is now a top priority, as it just cannot deliver adequate cluster position resolution (due to inability to properly handle overlapping clusters) in central Au+Au events with pileup.

Having said that, the present performance I see is:

Upsilon resolution:

-------------------------

  Low occupancy:  < 80 MeV

  Mb events with 50 kHz pileup:  102 +/- 4 MeV

  Central events with 50 kHz pileup: 150  +/- 12 MeV

These resolutions are inflated because we are using the field map in the simulations, while using a constant field in the acts track fitting. We need the Acts developers help to get the field map working - it is in their queue. Tests with simulations done in a constant field show that the resolutions will decrease by 5-10% when the simulation and fitter use the same field.

So the place where we have a problem is in 0-10% most central events. I think we will have to upgrade to a better clustering algorithm to get the resolution down to where we want it in those events.

Note that Sasha Lebedev reported to me recently that he sees 120 MeV Upsilon resolution for 0-10% central Hijing events with no pileup. I know that could be improved with different track cuts. So, certainly, turning down the collision rate would help. But it is way too early to think we will need to do that. I am confident that we will be able to reach 125 MeV in central collisions with 50 kHz pileup. I hope to do better.

Tracking efficiency:

--------------------------

  Low occupancy:  ~ 95%

  Mb events with 50 kHz pileup:  82% with 90-95% MVTX matching efficiency

  Central events with MB pileup: 52% with 90-95% MVTX matching efficiency

This is a problem, and it is a major focus now. We discussed some ideas of how to improve track seeding at the meeting this morning, and Michael is implementing them now. I am presently working on improving the TPC-silicon track matching. But, in the end, I think we will need better clustering to get really good tracking efficiency in central events.

In summary, tracking is very much a work in progress, and there are still many knobs to turn to improve performance in central events.

Cheers

Tony

---

From: sPHENIX-tracking-l <sphenix-tracking-l-bounces AT lists.bnl.gov> on behalf of Jamie Nagle <jamie.nagle AT colorado.edu>
Sent: Monday, February 8, 2021 9:39 PM
To: sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>
Subject: [Sphenix-tracking-l] Questions from a place of ignorance

 

Hello Tracking Experts,

I have been only somewhat following the progress on understanding track distortions, corrections, and momentum resolution and its relation to the TPOT proposal.    In looking back through recent presentations, I am missing part of the big picture -- which is certainly my fault.

I thought I would ask a few questions to see what quantitative answers exist.

* As a starting point, until September 2020,  the sPHENIX plan  was to have 150 kHz of AuAu  minimum bias collisions - a large fraction which would be outside the z  vertex, i.e. |z| < 10 cm.   For pp collisions the highest rates were 10 MHz or more, and now with the  plan only having pp 200 GeV and a crossing angle this is down to ~ 2 MHz.

* Only in September 2020 with  the new sPHENIX Beam Use Proposal did we give up  these high rates for calorimeter measurements only, and settle that the highest AuAu minimum bias rates would be approximately 50 kHz or so.  

* What was the original performance spec for the TPC to handle distortions at and achieve the benchmark Upsilon mass resolution?    

* Some of the current momentum resolution / upsilon resolution plots look like that program is lost (or never was).    It is possible to produce a figure of (1) charged particle momentum resolution versus p up to 50 GeV, and (2) upsilon 3-state mass resolution for the following equivalent rates: 1 kHz, 10 kHz, 25 kHz, 50 kHz.     At what rate do the effects of degraded resolution  really kick in?

* How much of the issue  with fluctuations in IBF and corrections is CPU time versus simply a  loss of information (i.e. impossible to recover)?     Where do the outer MMGs come into play here in both cases?

Some of this information may be critical in potentially re-thinking the run plan to make sure a "good" data set, though smaller, can be assured, while testing at higher rates.   Also, I thought there were ideas with regards to how small in radius we have active in the TPC that influences things -- is that now set in stone or still being discussed?    Same with the gain and gas question.

Thanks for any  help in this direction.

Sincerely,

Jamie

||------------------------------------------------------------------------------------------
|| James L. Nagle   
|| Professor of Physics, University of Colorado Boulder

|| EMAIL:   jamie.nagle AT colorado.edu
|| SKYPE:  jamie-nagle        
|| WEB:      http://spot.colorado.edu/~naglej 
||------------------------------------------------------------------------------------------